Based on the national migrant population dynamic monitoring survey data, the hierarchical linear model analysis is used to explore the influence of community characteristics on the social interaction of the migrant population. Studies have shown that: from the perspective of direct impact, the proportion of the migrant population in the community and the regional GDP per capita have a restraining effect on the social interaction, while community public services and public activities can help to improve the social interaction of the migrant population; from the "cross-layer" interaction effect’s results, the proportion of the migrant population in the community has strengthened the positive impact of "owner of the houses" while reducing the negative impact of "social exclusion". Community activities help to improve the social interaction of the "young" migrant population, but it further exacerbates the "age disadvantage" of the "older" migrants in social interaction. In addition, the increase in regional GDP per capita can improve the social interactions of the floating population with "low-end occupations". Research implications: Strengthening community governance and improving the community’s communication environment can effectively prevent the involution of the young migrant population’s communication and reduce the risk of community segmentation that may be caused by the migration of migrant population and family movement.